agents

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). The skill's docs and workflow explicitly allow defining webhook tools with arbitrary api_schema.url entries in conversation_config.agent.prompt.tools (references/client-tools.md and SKILL.md), and webhook responses are passed back to the agent and used in conversation, meaning untrusted third-party API/pages can be fetched and their content can influence the agent's decisions and next actions.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The embed script https://unpkg.com/@elevenlabs/convai-widget-embed is loaded at runtime via a tag and executes third‑party JavaScript in the client (a required dependency for the widget), so it is a runtime external dependency that executes remote code.